Improving Glassdoor's Give-to-Get Review Gate: Funnel Metrics and Experiments

Quick Overview

A funnel optimization question about Glassdoor's rule that visitors must submit a review before reading others. It tests whether a candidate can set the right goal and guardrails, map each funnel step to a metric, diagnose drop-offs by segment and design experiments that protect review quality.

Improving Glassdoor's Give-to-Get Review Gate: Funnel Metrics and Experiments

Company: Glassdoor

Role: Product Manager

Category: Analytics & Experimentation

Difficulty: medium

Interview Round: Other

On Glassdoor, a person who wants to read employer reviews must first contribute a review of their own. You are the product manager responsible for this requirement. How would you optimize the funnel it creates, from the moment a visitor runs into the requirement to the moment they are reading reviews? ```hint One gate, two jobs The requirement blocks readers and also produces the reviews everyone reads. Settle which outcome you are optimizing, and what must not get worse, before hunting for drop-off points. Then ask which visitors cannot pass the gate at all. ``` ### Clarifying Questions - What counts as a contribution: only a company review, or could a salary report or an interview experience also unlock access? - Is success more approved reviews, more visitors who reach the content, or more registered users who come back, and which wins when they conflict? - Does the requirement appear on the first page view or after some free content, and does it behave the same on the website and in the mobile app? - How is review quality judged today, and how many submissions fail moderation? ### What a Strong Answer Covers - A goal that balances contribution volume against reader experience and review quality, with explicit guardrails. - The funnel broken into concrete steps, each with a metric, and cut by segments such as traffic source, device and employment status. - Causes of drop-off at each step (nothing to review, fear of being identified, form effort, unclear value) and fixes matched to those causes. - An experiment plan that can detect harm to review quality and honesty, not just a rise in submissions. ### Follow-up Questions - Shortening the form raises submissions by 25 percent, but reviews get shorter and more extreme. Do you ship it? - How should a student who has never had a job get access to reviews? - How would you tell whether the requirement sends visitors to a competitor instead of converting them?

Overview: A funnel optimization question about Glassdoor's rule that visitors must submit a review before reading others. It tests whether a candidate can set the right goal and guardrails, map each funnel step to a metric, diagnose drop-offs by segment and design experiments that protect review quality.

Glassdoor logo
Glassdoor
Mar 17, 2025
mediumProduct ManagerOtherAnalytics & Experimentation
0
0

On Glassdoor, a person who wants to read employer reviews must first contribute a review of their own. You are the product manager responsible for this requirement. How would you optimize the funnel it creates, from the moment a visitor runs into the requirement to the moment they are reading reviews?

Clarifying Questions Guidance

  • What counts as a contribution: only a company review, or could a salary report or an interview experience also unlock access?
  • Is success more approved reviews, more visitors who reach the content, or more registered users who come back, and which wins when they conflict?
  • Does the requirement appear on the first page view or after some free content, and does it behave the same on the website and in the mobile app?
  • How is review quality judged today, and how many submissions fail moderation?

What a Strong Answer Covers Guidance

  • A goal that balances contribution volume against reader experience and review quality, with explicit guardrails.
  • The funnel broken into concrete steps, each with a metric, and cut by segments such as traffic source, device and employment status.
  • Causes of drop-off at each step (nothing to review, fear of being identified, form effort, unclear value) and fixes matched to those causes.
  • An experiment plan that can detect harm to review quality and honesty, not just a rise in submissions.

Follow-up Questions Guidance

  • Shortening the form raises submissions by 25 percent, but reviews get shorter and more extreme. Do you ship it?
  • How should a student who has never had a job get access to reviews?
  • How would you tell whether the requirement sends visitors to a competitor instead of converting them?